fix(definition): accept the legacy 'wildcard = {}' tool scope table by senamakel · Pull Request #441 · tinyhumansai/tinyagents

senamakel · 2026-10-13T23:06:37Z

Existing OpenHuman agent TOML spells an unrestricted scope as [tools] wildcard = {}. After #438 the untagged wire enum rejected it, so such a definition was silently skipped by the loader. ToolScope now parses with an explicit visitor accepting "wildcard", wildcard = {}, { named = [..] } and a bare list. Regression test: every_tool_scope_spelling_in_existing_toml_still_parses. Host follow-up: openhuman#7773.

Existing OpenHuman agent TOML spells an unrestricted scope as [tools] wildcard = {}. The untagged wire enum rejected it; parse the scope with an explicit visitor. Co-authored-by: Medulla <medulla@tinyhumans.ai>

coderabbitai · 2026-10-13T23:06:49Z

Warning

Review limit reached

Run on-demand review

_{This review includes 2 billable files and costs up to $0.50.}

Or wait 24 minutes for your next included review.

Check out review usage here.

View limit details

Limit details: You’ve used all 2 included reviews currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Organization UI
Review profile: CHILL
Plan: Advanced
Run ID: f0cc06df-4f69-47bd-a26c-76b9da1ae5b0

📥 Commits

Reviewing files that changed from the base of the PR and between f1e64cc and 254c7aa.

📒 Files selected for processing: 2

crates/tinyagents-definition/src/execution_spec.rs
crates/tinyagents-definition/src/execution_spec_tests.rs

Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

_{Comment @coderabbitai help to get the list of available commands.}

chatgpt-codex-connector · 2026-10-13T23:06:49Z

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-13T23:09:49.724367Z	`254c7aa`	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 254c7aa8d0

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-10-13T23:09:47Z

+                    .ok_or_else(|| A::Error::custom("empty tool scope table"))?;
+                let scope = match key.as_str() {
+                    "wildcard" => {
+                        map.next_value::<IgnoredAny>()?;


Reject arbitrary wildcard payloads

When an authored TOML/JSON scope uses the wildcard key with anything other than the legacy empty table (for example, wildcard = false or {"wildcard":["read"]}), IgnoredAny discards that payload and returns ToolScope::Wildcard. The previous enum deserializer rejected these malformed shapes, while Wildcard grants every parent tool and passes the loader's semantic diagnostics, so a typo or attempted restriction now fails open; deserialize and validate an empty map specifically instead of ignoring every value.

Useful? React with 👍 / 👎.

tinysweeper · 2026-10-14T02:32:43Z

Tiny Sweeper review

Tiny Sweeper completed its review; deterministic results follow.

State: Ready for maintainer review
Priority: none
Reviewed head: 254c7aa8d0e0
Updated: 2026-10-14T02:38:56Z

Review snapshot

Change surface	Files	Review signal
Production	0	Active findings
Tests	2	Noted findings
Documentation	0	Resolved findings
Configuration	0	Pending checks/questions

Completeness: Complete
Test assessment: No supported feature-to-test mapping was available; this does not mean tests are absent or passed.

What changed

`ToolScope` loses its derived untagged `Wire`/`Tagged` enum and gains a hand-written `Deserialize` visitor in `crates/tinyagents-definition/src/execution_spec.rs` that accepts the `"wildcard"` string, a sequence of tool names (bare list), a map with a single `named` key, and the legacy map form `wildcard = {}`. Unknown variants, an empty table, and a table with more than one key are rejected with errors. A supporting test was added to `crates/tinyagents-definition/src/execution_spec_tests.rs`.

Features

Modified — ToolScope deserialization accepts the legacy `wildcard = {}` table form: The hand-written visitor over `deserialize_any` now parses the legacy TOML table spelling `[tools] wildcard = {}` in addition to the previously accepted `"wildcard"` string, bare name list, and `{ named = [..] }` forms, restoring compatibility with existing on-disk definitions; unknown variants, an empty table, and a multi-key table still error out. (crates/tinyagents-definition/src/execution_spec.rs)
Added — Unit test pinning every accepted tool scope spelling: A new test parses each supported spelling through a `Holder` wrapper and asserts concrete expected `ToolScope` values, including `[tools] wildcard = {}`, `tools = "wildcard"`, `[tools] named = ["a", "b"]`, the bare list `tools = ["a"]`, and the empty `named = []` list, so a regression in any accepted spelling would fail a test. (crates/tinyagents-definition/src/execution_spec_tests.rs)

Tests

No supported feature-to-test mapping was produced. Test execution is not inferred.

Findings

No active actionable findings.

Before merge

None.

How this fits together

flowchart LR
  n0["ModelSpec<br/>changed"]:::changed
  n1["pin"]:::impacted
  n2["...reaches_the_resolver_without_an_exact_pin"]:::impacted
  n3["Hint"]:::impacted
  n4["deserialize"]:::impacted
  n1 -->|uses| n3
  n2 -->|uses| n0
  n2 -->|calls| n3
  n2 -->|tests| n3
  n4 -->|calls| n3
  n4 -->|uses| n3
  classDef changed fill:#0d4429,stroke:#238636,color:#e6edf3
  classDef impacted fill:#161b22,stroke:#6e7681,color:#c9d1d9
  classDef flagged fill:#5a1e02,stroke:#d93f0b,color:#ffffff
  classDef blocking fill:#67060c,stroke:#f85149,color:#ffffff

Agent review details

critique

Conclusion: Success
Scope reviewed: all assigned evidence
Lane summary: The rewrite of the `ToolScope` `Deserialize` impl replaces the derived untagged enum with a hand-written visitor over `deserialize_any`. The canonical forms in the new test (`"wildcard"`, bare list, `[tools] named = [..]`, `[tools] wildcard = {}`) are handled, but the previous implementation's `Tagged` enum was an externally-tagged enum accepting the *newtype variant* spellings `wildcard = {}` and `named = [...]`, so the set of accepted spellings appears preserved. Before concluding, I want to check two things I cannot see: how `ToolScope` is serialized for TOML (the `Serialize` derive with `rename_all = _The code index is behind this pull request (indexed at `d523fe4a8a8c`), so retrieved context may be out of date._

security

Conclusion: Success
Scope reviewed: all assigned evidence
Positive: Every branch of the visitor either produces `Wildcard`/`Named` or returns an error for an unknown variant, an empty table, or a multi-key table, so no malformed input can silently widen a tool scope.
Lane summary: The hand-written `Deserialize` for `ToolScope` replaces the previous untagged derive and accepts the same wire forms plus the legacy TOML table form (`[tools] wildcard = {}`). Every branch either produces `Wildcard`/`Named`, or returns an error for unknown variants, an empty table, or a multi-key table, so no malformed input can silently widen a tool scope. The new tests exercise each spelling. No security or correctness defect found in this diff. _The code index is behind this pull request (indexed at `d523fe4a8a8c`), so retrieved context may be out of date._

tests

Conclusion: Success
Scope reviewed: all assigned evidence
Positive: The added test pins every accepted spelling, including the legacy `wildcard = {}` table and the empty named list, against concrete expected values, so a regression in any of them fails a test.
Positive: The previously accepted forms are unchanged, and the new rejection branches reject exactly the inputs the old untagged enum also rejected, so no behaviour regresses silently.
Lane summary: The change replaces ToolScope's untagged deserialize with a visitor that also accepts the legacy TOML table form `wildcard = {}`, and adds `every_tool_scope_spelling_in_existing_toml_still_parses`, which pins every accepted spelling (`wildcard = {}`, the `"wildcard"` string, `named = [..]`, and the bare list, including the empty list) against concrete expected values — so a regression in any of them fails a test. The previously accepted forms are unchanged, and the new rejection branches (empty table, two keys) reject exactly the inputs the old untagged enum also rejected, so I found no behaviour that regresses silently and no defect to anchor. _The code index is behind this pull request (indexed at `d523fe4a8a8c`), so retrieved context may be out of date._

commits

Conclusion: Neutral
Scope reviewed: all assigned evidence
Positive: Commits lane found nothing sensitive in what this pull request commits.
Lane summary: Nothing sensitive found in what this pull request commits.

description

Conclusion: Success
Scope reviewed: all assigned evidence
Positive: The title and body match the code exactly: the previously accepted forms still parse and no repository rule is reported as broken.
Lane summary: The diff replaces ToolScope's untagged wire enum with a hand-written visitor that accepts `"wildcard"`, the legacy table `wildcard = {}`, `{ named = [..] }` and a bare name list, with a new unit test covering each spelling; the title and body match the code exactly, the previously accepted forms still parse, and no repository rule is broken. Nothing here looks unsafe to merge. _The code index is behind this pull request (indexed at `d523fe4a8a8c`), so retrieved context may be out of date._

Evidence and run details

Models: ladder/vectors-oai3, deepseek-v4.1-flash
Spend: $0.004330
Tokens: 193580 input · 60165 output · 17024 cached · 180 embedding

Head	State	Pass summary
`254c7aa8d0e0`	ready for maintainer review	0 active finding(s), 0 resolved finding(s) (at 2026-10-14T02:38:56Z)

tinysweeper

tinysweeper found nothing blocking. Approving.

             $0.0043 · 193,580 in / 60,165 out · 17,024 cached (9%) · ladder/vectors-oai3, deepseek-v4.1-flash · 180 embedded
critique:    $0.0001 · 58,587 in  / 706 out    · 6,144 cached (10%) · deepseek-v4.1-flash
security:    $0.0009 · 57,450 in  / 12,410 out · 3,456 cached (6%)  · deepseek-v4.1-flash
tests:       $0.0018 · 48,546 in  / 25,021 out · 2,816 cached (6%)  · deepseek-v4.1-flash
description: $0.0014 · 26,070 in  / 19,990 out · 4,608 cached (18%) · deepseek-v4.1-flash

fix(definition): accept the legacy 'wildcard = {}' tool scope table

254c7aa

Existing OpenHuman agent TOML spells an unrestricted scope as [tools] wildcard = {}. The untagged wire enum rejected it; parse the scope with an explicit visitor. Co-authored-by: Medulla <medulla@tinyhumans.ai>

chatgpt-codex-connector Bot reviewed Oct 13, 2026

View reviewed changes

senamakel merged commit 2e706e8 into tinyhumansai:main Oct 13, 2026
11 checks passed

senamakel mentioned this pull request Oct 14, 2026

[openhuman#7523] tinyagents: third pass, middleware, ladders, turn policy, graph tracing, dialect detector #444

Closed

10 tasks done

tinysweeper Bot approved these changes Oct 14, 2026

View reviewed changes

tinysweeper Bot added the priority: p3 Whenever. Cosmetic, a nicety, or a cleanup with no user visible effect. label Oct 14, 2026

Navigation Menu

Repository navigation

fix(definition): accept the legacy 'wildcard = {}' tool scope table - #441

fix(definition): accept the legacy 'wildcard = {}' tool scope table#441
senamakel merged 1 commit into
tinyhumansai:mainfrom
senamakel:tool-scope-legacy-wildcard

senamakel commented Oct 13, 2026

Uh oh!

coderabbitai Bot commented Oct 13, 2026

Review limit reached

Uh oh!

chatgpt-codex-connector Bot commented Oct 13, 2026 •

edited

Loading

Uh oh!

chatgpt-codex-connector Bot left a comment

Uh oh!

chatgpt-codex-connector Bot Oct 13, 2026

Uh oh!

Uh oh!

tinysweeper Bot commented Oct 14, 2026 •

edited

Loading

critique

security

tests

commits

description

Uh oh!

tinysweeper Bot left a comment

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

senamakel commented Oct 13, 2026

Uh oh!

coderabbitai Bot commented Oct 13, 2026

Review limit reached

Uh oh!

chatgpt-codex-connector Bot commented Oct 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codex Review Summary

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector Bot Oct 13, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

tinysweeper Bot commented Oct 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Tiny Sweeper review

Review snapshot

What changed

Features

Tests

Findings

Before merge

How this fits together

critique

security

tests

commits

description

Uh oh!

tinysweeper Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

chatgpt-codex-connector Bot commented Oct 13, 2026 •

edited

Loading

tinysweeper Bot commented Oct 14, 2026 •

edited

Loading